1. Field of the Invention
The present invention relates to an image processing technique for repetitively detecting, from image data, an object satisfying predetermined conditions and displaying a detection result of the object.
2. Description of the Related Art
An imaging apparatus can repetitively detect, from image data, an object satisfying predetermined conditions, for example, to improve the following situations discussed herein below.
For instance, a camera may have an auto-focus function for automatically adjusting focus on a target subject to be photographed. In general, the camera selects one or plural focus detection areas, and adjusts the lens position of its imaging optics system to focus on a subject in the selected focus detection area. Then, the camera performs exposure compensation processing by enlarging a weighting factor applied to a brightness value of the main subject located in the focus detection area.
However, the focus detection area can occupy a relatively limited area on the screen. When the main subject is present outside the focus detection area, it is difficult to focus on the main subject. Furthermore, even if a main subject is present in the focus detection area, the focus adjusting action may be erroneously applied to another object different from the target (i.e., main subject). For example, when a subject in a different focus detection area is positioned more adjacently to a camera than the main subject, the camera may regard a different subject in another focus detection area as a main subject and may erroneously apply a focus adjusting action to this subject.
To avoid this drawback, it is possible to request a user to instruct a focus detection area where a main subject is present each time the user shoots the target, although it is not convenient for the user. In view of the above, Japanese Patent Application Laid-open No. 2003-107335 discloses a camera that can automate the processes of detecting a main subject from obtained image data using a shape analysis, displaying a focus detection area corresponding to the detected main subject, and performing a focus adjusting action applied to the focus detection area.
According to the aforementioned camera, the image data is entirely searched to detect a main subject and accordingly the focus adjusting action can be applied to the main subject wherever the main subject is present in an object field. Furthermore, to momentarily track the main subject, the detecting action of the main subject based on the shape analysis must be performed periodically.
However, an apparatus capable of automatically detecting a main subject may erroneously select a subject that a user does not intend to shoot. Hence, it is necessary to let a user confirm a main subject detected by the camera.
Furthermore, a liquid crystal monitor or other display unit can continuously display image data to let a user observe the movement of a subject. In such a case, it is desirable to update the detection result of a main subject in accordance with a detected movement of the main subject. To this end, the processing for updating the detection result of a main subject should be performed periodically. And, the latest region where the main subject is present should be continuously displayed.
More specifically, when a main subject is detected, a frame indicating the position of a detected subject can be superimposed on an image captured by the camera. In this case, as a practical method for detecting a main subject, it is possible to detect a front or oblique face of a subject person based on the spatial relationship between both eyes, a nose, and a mouth on a face (refer to Japanese Patent Application Laid-open No. 2002-251380).
However, according to the above-described detection method, if a subject person blinks and closes an eye or suddenly turns his/her face to look away, one eye will not be recognized on a captured image and accordingly the camera will fail in detecting a main subject. As a result, as shown in FIG. 14, the frame indicating the detected region will temporarily disappear from the monitor screen, whereas the main subject remains at the same place.
For example, posing for a while is hard for a child who is waiting for completion of a shot. Thus, if the above-described method is used to detect a child, the camera may temporarily fail in detecting a main subject. Such drawbacks will induce an undesirable phenomenon repeating the display and erasure of the frame indicating the position of a main subject in a short period of time. In this situation, the image displayed on a display unit will be unstable and a user will be unable to surely observe the movement of a subject.
Similar problems will commonly arise when a main subject is repetitively detected from a moving image or from continuously changing image data and a detection result is displayed on a display unit.
Also it is noted that the above-described problems are not limited to cameras or other image capturing devices. For example, similar problems will arise in application software that can detect a target object from transferred moving image data and display a detection result.
Therefore it would be desirable to provide an apparatus which has a function for repetitively detecting, from image data, an object satisfying predetermined conditions, and of which further can stably display a detection result even when a target object cannot be temporarily detected